Inheritance

18 min

Explanation

Inheritance lets a class (the subclass) reuse and extend another class's (the parent's) behavior, instead of rewriting it from scratch. class Rectangle(Shape) means "a Rectangle IS-A Shape, plus whatever extra Rectangle defines."

class Shape:
    def area(self):
        raise NotImplementedError   # every subclass MUST override this

class Rectangle(Shape):
    def __init__(self, w, h):
        self.w = w
        self.h = h

    def area(self):
        return self.w * self.h

Shape.area deliberately raises an error — it's a placeholder saying "every real shape must define its own area," a pattern called an abstract method.

Try it

isinstance() checks the whole inheritance chain, not just the exact class -- a Rectangle passes as both Rectangle AND Shape, since inheritance is an IS-A relationship.

Loading editor…

Explanation

When a subclass needs the parent's setup logic PLUS a bit more, calling super().__init__(...) runs the parent's __init__ without duplicating its code:

class Square(Rectangle):
    def __init__(self, side):
        super().__init__(side, side)   # a square is a rectangle with w == h

Square gets area() for free — it never defines its own, so Python looks up the inheritance chain and finds Rectangle.area, which works perfectly since Square has self.w and self.h set (by the super() call) exactly like any other Rectangle would.

Exercise

Given `Shape` (base class) and `Rectangle(Shape)` below, implement `Rectangle.area(self)`, returning `width * height`.

Exercise

Given `Square(Rectangle)` below, implement `Square.__init__(self, side)` by calling `super().__init__(side, side)` — reuse Rectangle's setup instead of duplicating it.

Quiz

What does calling super().__init__(...) inside a subclass's __init__ do?

Checkpoint

You can define a subclass that overrides a parent method, and use super() to reuse a parent's init instead of duplicating its logic.